The Value Of Panels In Modeling Big Data
Hardly a day goes by without an industry report on audience fragmentation. Of course, it’s not a new phenomenon. With the rise of cable in the 80s, digital broadcast satellite in the 90s, Internet video in the 2000s and more recently over-the-top options, television audiences have enjoyed a steady stream of new programming choices year after year: More networks, more niche programs, and more ways to watch them.
For the research community however, that increased diversity has come at a price, and the accelerating pace of change in recent years is straining the panel-based measurement capabilities that the industry has historically relied on to monitor viewing activity. It’s simply become a challenge to assemble panels large enough to provide stable measurement for programs with small audiences.
Return Path Data (RPD) represents an opportunity to overcome that problem, but only if the limitations and biases in these data can be corrected and validated. This paper describes how panels can effectively correct for these limitations and help validate the ratings derived from RPD datasets.
Panels and RPD together represent a winning combination for accurate and stable video audience measurement.